Brian,
> -----Original Message-----
> From: [email protected]
> [mailto:[email protected]] On Behalf Of
> Brian P. Poi
> Sent: Tuesday, September 02, 2008 5:48 PM
> To: [email protected]
> Subject: RE: st: RE: clustering option after ivprobit
>
> On Tue, 2 Sep 2008, Schaffer, Mark E wrote:
>
> > Brian,
> >
> >
> > Just out of curiosity - why isn't cluster-robust (or just
> robust, for
> > that matter) allowed after two-step? I've noticed this is also the
> > case for the related estimators that also have two-step versions,
> > -heckman- and -treatreg-.
> >
> > --Mark
> >
>
> Mark,
>
> We simply haven't worked out the details of the robust
> variants of the VCEs for the two-step estimators. Hardin
> (2002, Stata Journal) shows the general procedure to obtain
> the robust variance estimator for two-step models:
>
> http://www.stata-journal.com/article.html?article=st0018
>
>
> I'd imagine a similar approach would work for the
> cluster-robust version as well.
Thanks, Brian. I had mistakenly thought the absence of this option for
two-step estimators in Stata was due to something fundamental, but I
couldn't think of what it was.
I'm sure your agenda is as full as always, but I think it would be
useful to add these in. Two-step estimators are often very handy in
practice. Sometimes getting ML versions to converge is just impossible,
or it's possible but it take so long that trying different
specifications is painful.
Cheers,
Mark
> -- Brian Poi
> -- [email protected]
> *
> * For searches and help try:
> * http://www.stata.com/help.cgi?search
> * http://www.stata.com/support/statalist/faq
> * http://www.ats.ucla.edu/stat/stata/
>
--
Heriot-Watt University is a Scottish charity
registered under charity number SC000278.
*
* For searches and help try:
* http://www.stata.com/help.cgi?search
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
